Costs and consequences in perioperative care: Analytic models in studies on pain treatment and on haemodynamic optimization of elderly patients

Sammanfattning: ABSTRACT Background Because resources are scarce in health care, costs and consequences of new interventions must be assessed to support informed policy decisions. This thesis analyses the cost-effectiveness of advanced postoperative pain treatment and perioperative haemodynamic optimization by applying decision modelling as an analytic framework. 1. Postoperative pain treatment refers to epidural analgesia and to patient-controlled in travenous analgesia. Based on the superior analgesic effect found in clinical trials, epidural analgesia is regarded as the gold standard following major surgery, but a drawback is the high failure rate (10–15%). Considering that approximately 40 000 patients are treated by epidural analgesia per year in Sweden, costs and consequences of this clinical problem are substantial. 2. Haemodynamic optimization refers to fluid protocols targeted to increase blood flow, referred to as goal-directed haemodynamic treatment. These protocols are beneficial in the perioperative care of high-risk patients, but there is lack of evidence in elderly patients. In Sweden 20 000 patients are operated on each year for proximal femoral fracture, with poor postoperative outcome. Large trials are required to assess whether any protocol of the goaldirected haemodynamic treatment is beneficial in the elderly population, in terms of outcome and health care costs. Considering the cost and complexity of such a trial, a prior costeffectiveness analysis might be adequate to guide the initiation of such a trial. Methods 1. Epidural analgesia vs. patient-controlled intravenous analgesia: Paper I: A decision-analytic cost-effectiveness model was developed to analyse data of a clinical database on pain treatment following major abdominal surgery. Paper II: Postoperative intensive care costs were analysed on data from patients included in a previously published trial on postoperative pain treatment following thoracoabdominal oesophagectomy. 2. Goal-directed haemodynamic treatment vs. traditional fluid treatment in elderly patients: Paper III: A decision-analytic cost-effectiveness model was developed, and relevant data from published trials and national registries were analysed. As the clinical outcome for elderly patients was previously unknown, reasonable estimates are applied in the model. Paper IV: The prior cost-effectiveness analysis (Paper III) guided the initiation of a large (n = 460) randomized clinical trial in elderly patients with proximal femoral fracture, and interim analyses of safety and efficacy were conducted (n = 100). Given the interim efficacy data, the monetary value of further data collection was analysed by calculating the expected value of perfect information. Results 1. The epidural analgesia is not cost-effective and no saving of the postoperative costs can be achieved, given the available evidence in Swedish clinical routine (Papers I–II). 2. The goal-directed haemodynamic treatment is predicted to be cost-effective in elderly patients, based on the available evidence and on the prior estimates of clinical outcome before the initiation of the trial. The expected value of perfect information is high, indicating that collecting further data by continuing the trial is potentially worthwhile (Papers III–IV). Conclusions 1. The analyses of epidural analgesia challenge its position as the gold standard and may assist revision of clinical policy decisions on postoperative pain treatment. 2. The analyses of the goal-directed haemodynamic treatment in elderly patients using a decision-analytic cost-effectiveness model suggest the usefulness of the initiation and continuation of a large clinical trial.